Oh, I see. This is caused by me mixing your script and manday's scripts together in my head I think.
You're detecting if the model intersects a line at various altitudes, right? Hence the "_alts = [1.4, 1, 0.2];". This still means that gigantous and tiny models (there where some of those for ofp that I just couldn't stop lauging at ever) will bug but in general it's not a big issue for most purposes. Is that understanding correct?
Now, manday's *does* return a floating point indicator of the unit's "spine2". Meaning that if the unit is made differently then the bis units, it's bounds may change to the unexpected. Is that understanding correct?
If both of those are correct understandings, then clearly your method is far superior.